How to Untangle an Argument

I’m studying Think Again I: How to Understand Arguments on Coursera. These are my revision notes for weeks 2. My notes for the the previous week are at: How to Spot an Argument.

How to Identify an Argument

Argument Markers

Definition of an argument:

  • a series of sentences, statements or propositions,
  • where some are the premises
  • and there is one conclusion
  • where the premises are intended to give a reason for the conclusion
  • When are certain propositions given as reasons for other propositions?
    • There are certain words which show this. These are called “argument markers
  • “so”, “therefore”, “thus”, “hence”, “accordingly”
    • Indicates that one proposition is a reason for another.
    • e.g. “I am tall, and I am good at sports” versus “I am tall, so I am good at sports”
    • Switching the order of the prepositions in the first sentence does not change the meaning of the sentence, but the order can’t change in the second without changing the meaning.
  • Kinds of argument markers
    • Conclusion markers: “so”, “therefore”, “thus”, “accordingly”
      • indicates that the proposition directly after is a conclusion
    • Reason markers: “because”, “for”, “as”, “for the reason that”, “for the reason why”, “since”
      • indicates that the proposition directly after is a premise or reason
    • Context matters. Some works in English have multiple meanings.
      • e.g. “since” could be a reason marker, or just a temporal signal (e.g. “It’s been raining since 7am”
      • e.g. “so” could be a conclusion marker, or not (e.g. You don’t have to eat so much”)
      • To test if the word actually is an argument marker, try this test: substitute a different argument marker (either conclusion or reason marker, depending on the sentence). e.g. “He’s so cool” => “He’s because cool” does not make sense, therefor “so” is not acting as an argument marker.
    • Conditionals: “if”, “then” are not argument markers, they just show what-if type of statements
      • If <antecedent> then <consequent>
      • e.g. “If I am rich enough, then I can buy a basketball team”. The proposition “I am rich enough” is not a true statement, and so this is not an argument

Standard Form

  • The order “because” propositions does not change the meaning of a sentence
    • e.g. “Because I am a professor, I teach classes” is the same as “I teach classes because I am a professor”
  • Contrast to “so”, which changes the argument
    • e.g. “I teach classes, so I am a professor”
  • To check, use the standard form.

   (1) Premise

(2) Premise


∴ (3) Conclusion (from 1-2)

Stopping the Regress

A Problem for Arguments

  • Premises are intended to be reasons for a conclusion
  • How do we understand if a person actually succeeds in providing valid reasons for the conclusion?
  • Justification:
    • An argument cannot justify you in believing that the conclusion is true unless you are justified in believing that the premises are true.
      • e.g. Life on Mars? Take this argument:
        • (1) There is at least on bacterium on Mars

        • ∴ (2) There is life on Mars
      • How do we know that the premise (1) is true? If we’re only guessing, then that is an invalid argument.
  • The problem of the sceptical regress:
    • If the premise is not justified, then we need to make an extra argument to justify that, such that the conclusion of the second argument is the premise of the first argument.
    • The second argument is also made up of premises, which also needs to be proven with another argument, and so on.
    • This chain can continue indefinitely. How do we resolve this?
  • Invalid solutions to the problem of sceptical regress:
    1. Start with a premise that is unjustified
      • Not a great solution, as it essentially allows you to prove anything, because you don’t have to justify antecedents
    2. Use an argument with a circular structure
      • This is no good! Essentially, the degenerate case this is an argument using its own conclusion as a premise, which is not allowed.
      • This means that a circular argument could be allowed to prove anything!
    3. Use an infinite chain of arguments
      • This doesn’t work, as we can keep on adding extra arguments to prove an unfounded premise. e.g. “There are two bacterium on Mars, therefore this is one bacterium on Mars”
  • Tricks for dealing with Sceptical Regress:
    • Assure the audience
      • Use a shared assumption that everyone agrees is true (e.g. “Honda’s are reliable”)
    • Guard your claim
      • Use an appeal to authority, as say “they’re probably right”
    • Discount objections
      • Use an appeal to authority, and give reasons why this authority is believable
    • Find shared assumptions to get the argument going
    • If you’re dealing with an audience that shares assumptions, it is easy
    • If you’re dealing with an audience that doesn’t share any of the assumptions, it’s much harder. You can use guards to move beyond the initial rejection of a premise.

Assuring

  • Assurance works if people trust the person making the argument. If they don’t trust the person, then this won’t work
    • e.g. “I assure you that smoking is bad for you health”.
    • This does not provide reasons for this premise, though it implies that you do have evidence.
  • Types of Assurances:
    • Authoritative
      • Cite an authority
      • e.g. “I assure you that smoking is bad for your health. The surgeon general has shown that it is bad for your health”
      • What is an authority?
        • If you know who the authority is and trust them, it is not a problem
        • However, if you don’t know who the authority is, you can’t really trust them (e.g. Watergate scandal, Deep Throat)
        • You have to be careful when people start citing an authority when the person isn’t actually an authority
        • Sometimes, authorities do make mistakes
    • Reflexive
      • Talking about yourself, e.g.
        • “I believe that…”
        • “I know that…”
        • “I am certain that…”
        • “I feel sure that…”
      • This works because people don’t like to question why other people feel sure about something. It can be impolite to question someone.
      • The fact that someone has “thought about this for years” doesn’t mean that they are actually right, though.
    • Abusive
      • They abuse you to get you agree with them by making a conditional abuse that only applies if you disagree with them. e.g.
        • “Nobody but a fool would believe that”
        • “Everybody believes this”
      • Appeal to common sense
        • This is a more subtle form of an abusive assurance: “It’s just common sense that…”
  • Benefits of Assurances:
    1. They save you time
      • We’ve got limited time.
      • You can’t look into every issue
    2. They help you avoid the sceptical regress problem
      • They let us move on with an argument,
  • Tricks with Assurances to Fool an Audience:
    1. Citations of untrustworthy authorities
    2. Distractions
      • It’s worth looking carefully when people make statements like this, as they may be trying to distract you from looking closely at an invalid premise:
        • “That’s obvious”
        • “It’s certain”
        • “I’m sure”
    3. Dropping assurances
      • Dropping the assurances can be illegitimate, for example when reporting goes like:
        • “He says…”
        • “It is reported that…”
        • “Sources have said that…”
        • “There are…”
  • When are assurances appropriate?
    1. Someone might question the assurance
    2. The audience accepts the authority
    3. It would be too much trouble to cite all the evidence
  • When are assurances not appropriate?
    1. No one would question the claim, anyway
    2. The authority is not trustworthy
    3. You are able to easily give the full explanation

Guarding

  • Guarding involves making the premises weaker so that it is harder to object to them e.g.
    • “We should not build any new nuclear power plans because they will explode”
    • It can be argued that the premise is unjustified (“Any new nuclear power plant will explode”)
    • The premise can be weakened like this:
      • “We should not build any new nuclear power plans because some of them might explode.”
    • An even weaker form:
      • “We should not build any new nuclear power plans because I believe that some of them might explode.”
  • When you see someone using guarding, you need to ask:
    • Why did they put in the guard?
    • Have they weakened the premise so much that the conclusion no longer follows?
  • Types of Guarding:
    • Extent
      • Weaken a premise by reducing the extent beyond what would otherwise be expected in the context
      • e.g. “All” -> “Most” -> “Many” -> “Some”
    • Probability
      • You can make a premise more defensible by showing that there are fewer ways for the premise to be wrong
      • e.g. “Absolutely certain” -> “probably” -> “likely” -> “there is a chance that” -> “might”
    • Mental
      • It is to do with the mental state of the person making the premise
      • e.g. “I know” -> “I believe” -> “I tend to believe” -> “I’m inclined to believe”

Discounting

  • Discounting is citing the possible objection that you think other people might be thinking of, in order to head it off by providing a quick and dirty response to it right there.
  • Functions of discounting terms:
    1. They assert two claims
    2. They contrast the two claims
    3. They emphasise one of the claims
  • e.g. “The ring is expensive, but it is beautiful”
    1. It asserts (or admits) that the ring is expensive
    2. It also asserts that the ring is beautiful
    3. It contrasts the two claims
    4. It emphasises the 2nd claim about beauty
  • Discounting terms:
    • “but”, “although”, “even if”, “even though”, “whereas”, “nevertheless”, “nonetheless”, “still”
    • “but” is similar to “and”, but it is different
      • the sentences on either side of “and” are reversible
      • the sentences on either side of “but” are not reversible
      • The word “but” indicates that the 2nd sentence is more important than the first
    • “although” de-emphases the sentence it is next to
  • The words above are not always used as discounting terms
    • Use the substitution method to test for what the function of the word is
    • e.g “He is sitting still” => “He is sitting although” (does not make sense, thus “still” is not a discounting term in this case)
  • The trick of discounting straw people
    • The arguer discounts easy objections to make people forget to think about the more difficult objections.
  • Another trick:
    • Arguers can combine the trick of discounting straw people with misuses of guarding and asserting
  • One rule of thumb:
    • Think about the objections that the arguer is not considering

Evaluation

  • e.g. “You ought to support my health care plan because it is good for the country”
  • Expressing preferences is not the same as making an evaluation
  • Evaluative language is all about the standards
    • “good” => meets the standards
    • “bad” => violates the standards
  • One trick:
    • when we call something “good”, we don’t specify the standard
    • this makes our claim more defensible since our standards cannot be questioned if they are not presented
  • Another trick:
    • we don’t have to agree on what the standards are
    • e.g. “fastest route” versus “most beautiful route”
  • Levels of evaluation
    • General: evaluation applies to standards or wide ranging rules, e.g.
      • “good” or “bad”
      • “ought” or “ought not”
      • “should” or “should not”
      • “right” or “wrong”
    • Specific: evaluation applies to individual items or a limited range of things, e.g.
      • “beautiful” or “ugly”
      • “cruel” or “kind”
      • “brave” or “cowardly”
      • “comfortable” or “uncomfortable”
    • Specific words are evaluative because of their connection to the more general evaluative words, e.g.
      • “beautiful” => “good”, “ugly” => “bad”
  • Contextually evaluative:
    • A word that can be either “good” or “bad” depending upon who hears it.
    • e.g. “liberal”
      • Republican: bad
      • Democrat: good
  • We will call language “evaluative” only when it is openly and literally evaluative, not when it is only contextually evaluative.
  • Combining positive and negative evaluative words can lessen heighten the level of the evaluation, e.g.
    • “good” => good
    • “pretty good” => Not as good as “good”
    • “pretty darn good” => very good
  • “too”: turns a positive or neutral evaluation bad, e.g.
    • This food is spicy => good
    • This food is too spicy => bad
  • Slanting
    • “slanting” is using evaluative terms without having reasons for the evaluation
    • slanting can be used to try and paper over weak arguments

Close Analysis

So far, we have covered the different roles language can play in an argument:

  • reason markers (R)
  • conclusion markers (C)
  • assuring terms (A)
  • guarding terms (G)
  • discounting terms (D)
  • evaluative terms (E+ or E-)

How do we bring all this together? The answer is close analysis.

In close analysis, you simply write out a passage, and mark all the words in the passage that fulfil the above roles.

Example passage for analysis, an op-ed by Robert Redford:

A Piece of “God’s Handiwork”, by Robert Redford

The Washington Post, 25th November 1997

Just over a year ago, President Clinton created the Grand Staircase Escalante National Monument to protect once and for all some of Utah’s extraordinary red rock canyon country. In response to plans of the Dutch company Andalex to mine coal on the Kaiparowits Plateau, President Clinton used his authority under the Antiquities Act to establish the new monument, setting aside for protection what he described as “some of the most remarkable land in the world.” I couldn’t agree more: For over two decades, many have fought battle after battle to keep mining conglomerates from despoiling the unique treasures of this stunning red rock canyon country. Now, we thought at least some of it was safe.

Not so. Shocking as it sounds, Clinton’s Bureau of Land Management (BLM) has approved oil drilling within the monument. BLM has given Conoco Inc., a subsidiary of the corporate giant DuPont, permission to drill for oil and gas in the heart of the new monument. You may wonder, as I do, how can this happen? Wasn’t the whole purpose of creating the monument to preserve it’s colorful cliffs, sweeping arches and other extraordinary resources from large scale mineral development? Didn’t the president say that he was saving these lands from mining companies for our children and grandchildren?

  • First paragraph
    • First sentence “Just over a year ago, President Clinton created the Grand Staircase Escalante National Monument to protect once and for all some of Utah’s extraordinary red rock canyon country.“:
      • “just” (G). It’s not evaluative (justice). Rather, it’s an indication of time (slightly more than a year ago). It could be classified as a guarding term to hold of arguments about exactly when the statement refers.
      • “to” (R). The “to” is actually quite significant. It’s some sort of argument marker. We can see that the next premise is talking about protecting the land. It’s a conclusion, so the “to” is acting as a reason marker
      • “protect” (E+). The word is indicating that it is a good thing to keep the land safe
      • “once and for all” (N = neutral): looks important, but doesn’t actually add anything to the argument. Rather, it’s an absolute statement about the duration of the protection (i.e. not guarded)
      • “some” (G). This is a guard term, as it reduces the scope of the protection
      • “extraordinary” (N)
    • Second sentence “In response to plans of the Dutch company Andalex to mine coal on the Kaiparowits Plateau, President Clinton used his authority under the Antiquities Act to establish the new monument, setting aside for protection what he described as “some of the most remarkable land in the world.”“:
      • “In response to” (R). It shows that the conclusion (monument created) was due to the following reason (plans of the Dutch company….)
      • “authority” (?). Tricky word to analyse, as it could be many things. Discounting term? Positive evaluation? Argument marker?
      • “under” (R). This shows the antiquities act is the reason why the President could do what he did
      • “to” (R). Also shows the reason
      • “protection” (E+)
      • Quotation marks “” (A). Assuring Clinton
    • Third sentence “I couldn’t agree more: For over two decades, many have fought battle after battle to keep mining conglomerates from despoiling the unique treasures of this stunning red rock canyon country.”:
      • “I couldn’t agree more” (A). Robert is increasing the assurance provided by Clinton
      • “For” (N). In this case, it’s not an argument marker, it’s just showing time
      • “many” (N). Not a guarding term in this case.
      • “battle” (E-). Battles are not a good thing
      • “to” (R). Shows the premise of why the battles are occuring
      • “despoiling” (E-)
      • “treasures” (E+)
    • Fourth sentence “Now, we thought at least some of it was safe.”:
      • “thought”, “at least” (G). Guarding
  • Second paragraph:
    • First sentence “Not so.”:
      • “so” (N). It’s not actually introducing a conclusion here
    • Second sentence “Shocking as it sounds, Clinton’s Bureau of Land Management (BLM) has approved oil drilling within the monument.”:
      • “shocking” (N). Robert thinks it’s bad, but the word itself is not intrinsically evaluative.
      • “as it sounds” (G). Guarding the “shocking” (it seems shocking to me)
    • Third sentence “BLM has given Conoco Inc., a subsidiary of the corporate giant DuPont, permission to drill for oil and gas in the heart of the new monument.”:
      • “giant” (N). Possibly intended to be evaluative, but giant itself is neutral
      • “permission” (E). Shows that the premise is not forbidden, but doesn’t actually indicate if that is good or bad
      • “heart” (N). Metaphor to invoke emotions
    • Fourth sentence “You may wonder, as I do, how can this happen?”:
      • “may” (G)
      • “how can this happen?”. Rhetorical question. Can cause the reader to provide an answer themselves (even unconsciously), and so have a stronger attachment to the argument
    • Fifth sentence “Wasn’t the whole purpose of creating the monument to preserve it’s colorful cliffs, sweeping arches and other extraordinary resources from large scale mineral development?”:
      • Once again, the sentence is rhetorical
      • “the whole purpose” (C). The conclusion
      • “to” (R). The premise
      • “preserve” (E+).
      • “colourful” (N)
      • “sweeping” (N)
      • “extraordinary” (N)
      • “resources” (E+)
    • Sixth sentence “Didn’t the president say that he was saving these lands from mining companies for our children and grandchildren?”:
      • Once again, this sentence is rhetorical
      • “Didn’t the president say” (A). Assurance from the president
      • “saving” (E+)
      • “for” (R)

 

 

One thought on “How to Untangle an Argument

Leave a comment